Rev. bras. oceanogr., 45(1/2):35-43, 1997

BENTHIC COMMUNITIES OF SHALLOW-WATER REEFS OF ABROLHOS,

Roberto Villaça1 & Fábio B. Pitombo2

lInstituto de Biologia da UniversidadeFederal Fluminense Departamento de Biologia Marinha (Caixa Postal 100644, 24001-970 Niterói, RJ, Brazil)

2Instituto de Biologia da Universidade Federal Rural do Rio de Janeiro (23851-970 Seropédica, RJ, Brazil)

. Abstract: The benthic communities of fringing and mushroom-shaped shalIow-water reefs of the Abrolhos region (southern coast of ) were surveyed. Line transects were used to estimate coral and alga1 percentage cover. Mussismilia braziliensis is the most conspicuous coral species in the majority of the communities surveyed, but turf algae make up the dominant cover in alI but one studied reef. In general, communities on mushroom-shaped reefs have higher diversity and higher coral cover than on fringing reefs. For both reef morphologies, the coral to alga cover ratio does not show marked differences between annual surveys, despite the high productivity characteristic ofthe dominant algal species.

. Resumo: Foram estudadas comunidades bentônicas de pouca profundidade nos recifes em franja e chapeirões da região de Abrolhos (costa sul da Bahia). A cobertura das espécies foi estimada pelo método de transect de linha. Mussismilia braziliensis é a espécie de coral mais importante na maioria das comunidades estudadas, porém as algas em tufo são os organismos dominantes em todos os recifes estudados à exceção de apenas um. Em geral, as comunidades nos chapeirões têm maior diversidade e maior cobertura de coral do que as dos recifes em franja. Nos dois tipos de recife, a relação de cobertura coral-alga não apresenta diferença significativa no tempo, apesar da alta produtividade característica do tipo de alga dominante.

. Descriptors: Benthic communities, Coral reefs, Abrolhos, Brazil.

. Descritores: Comunidades bentônicas, Recifes de coral, Abrolhos, Brasil.

Introduction global environmental impacts and the need for their conservation, a monitoring project was introduced to The Abrolhos complex, located off study the shalIow benthic communities of these reefs. the Eastem coast of Brazil, is the southemmost coral This study is part of a larger monitoring program reefs in the Atlantic Oceano The Abrolhos reefs coordinated by the Marine Park Direction to provide exhibit several different morphologies (Leão et aI., reef conservation. 1985, 1988), and are mainly characterized by a low In the present study, permanent transects were coral diversity with a high degree of endemism instalIed for annual monitoring of algal and coral (Laborel, 1969a, 1969b; Pitombo et a!., 1988) what cover between morphologicalIy different reefs with makes this an unique environment. the aim of contributing to the understanding of the Despite the ecological importance of this community structure of the southemmost coral reefs ecosystem, very little is known about its community in the Atlantic 0cêim. structure. Part ofthe reef area of Abrolhos lies within the first Brazilian Marine National Park, "Parque Study area Nacional Marinho dos Abrolhos" (Gonchorosky et aI. 1989) created in 1983. Because of the increased The Abrolhos reefs are spread over an concern about the susceptibility of reef ecosystems to approximate area of6.0000 km2 offthe south coast of the Bahia State (17°20' to 18°1O'S and 38°05' to 36 Rev. bras. oceanogr., 45(112), 1997

39°00'W) (Fig. 1). The reefs are grouped into two transects is shown in Table 1, along with the used ares: the first is found nearshore (10 to 20 kilometers eodes and depths. off the coastline) -Timbebas Reef, Parcel das Paredes Line transect methodology (Loya, 1972) was Reefs, Coroa Vermelha Reef, Sebastião Gomes Reef, used to describe quantitatively the structure of the and Viçosa Reef - and the seeond lies 60 to 70 reef community. Line transects of 10m long were kilometers offshore, surrounding the Abrolhos placed in characteristic zones of different reefs. A Arehipelago (Parcel dos Abrolhos Reefs) (Leão et aI., metal pin was fixed in each survey site so that the 1985, 1988). The prevailing winds blow from the same transeet could be repeated over time. northeast and east during most of the year with The folIowing community indices were used: speeds ranging from 4 to 15 knots. During the winter, southern winds are common with speeds 1 -Relative frequeney index (Fr =Fi / Ft X 100); ranging from 17 to 21 knots. Around the nearshore where: Fi = no. oftransectsin whicha speciesoccurs reefs, the depth of the shelf does not exceed 15 and Ft = total number oftransects) meters but the Parcel dos Abrolhos Reefs area, lie 2 - Taxonomic dominanee Dr = Di / 160 X 100; between 20 and 35 m deep. The water temperature where Di= total percentage eover of one taxon in alI varies from 24°C (winter) to 27°C (summer) and the transects and "160" is the total length in meters of tidal amplitude ranges from -0.1 to 2.4 m throughout alI transects. the year (Coutinho et aI., 1993). 3 - Index ofImportance (I = Fr X Dt /100) 4 - Shannon diversity (H') and Pielou's eveness (E)(Legendre & Legendre, 1979) Material and methods 5 - Functional group dominance (DG = ~Di X 100; where ~Di is the sum of the cover of each taxon that Sampling was carried out in three periods: belongs to a Functional group (ex. Coral, Zoanthids, January of 1992, December of 1992 and November of Turf algae, Frondose algae and Crustose coralline 1993. Five samplings sites were deftned, three of algae) in each transect. them were surveyed more than once. A list of

20' 10' 39" 50' 40'

\ 20' \ O Timbebas )\éJ Reef 30' .GuaritaI. -1 (\ nta Borbora I. 40' ~:l. ~ c~qva~ j~ ~s RedondaI. 50' ;;'? ~~':r" 3 V Abrolhos ' .2~' '" ~ . . ~ . Abrolhos O ArChipela~ I eu Coroa ~.V Siriba I. Archipelago \ ' 18°~~~ Vermelha ~ 5 I Reef Reef AbrolhosParcel dO!efsRee ~

I

10' ~ I 10 Km I lKm SuesteI. I ~I "- ",

Fig. 1. Map of the Abrolhos region, with the location of the studied reefs. (1) Fringing reef at Santa Barbara Island; (2) Fringing reef at Siriba Island; (3) Mushroom-shaped reef at the Parcel dos Abrolhos Reefs; (4) Mushroom-shaped reef at the southem border ofthe Timbebas Reef; (5) Mushroom-shaped reef dose to Sueste Island ofthe Abrolhos Archipelago. Vll.,LAÇA& PITOMBO: Benthic communities of Abrolhos 37

Table 1. Transect location, code used (code), depth (meters) and date of sampling

LOCATION CODE DEPTH MONTH Timbebas Reef TIN3 4 Nov 1993 Sueste Reef SUN3 6 Nov 1993 Parcel dos Abrolhos Reef ROD2 8 Dec 1992 Parcel dos Abrolhos Reef RON3 8 Nov 1993 Santa Barbara I. Coral Zone BCJ2 2.8 Jan 1992 Santa Barbara I. Coral Zone BCD2 2.8 Dec 1992 Santa Barbara I. Coral Zone BCN3 2.8 Nov 1993 Santa Barbara I. Palythoa Zone BPJ2 0.9 Jan 1992 Santa Barbara I. Palythoa Zone BPD2 0.9 Dec 1992 Santa Barbara I. Palythoa Zone BPN3 0.9 Nov 1992 Siriba Island Coral Zone SCJ2 1.9 Jan 1992 Siriba Island Coral Zone SCD2 1.9 Dec 1992 Siriba Island Coral Zone SCN3 1.9 Nov 1993 Siriba Island AIgae Zone SAJ2 1.3 Jan 1992 Siriba Island Algae Zone SAD2 1.3 Dec 1992 Siriba Island AIgae Zone SAN3 1.3 Nov 1993

The similarity between samples was estimated Phaeophyta). The 10m transects were placed at the using the Serensen index (S = 2a/b+c, where a = Palythol.\and in the coral zones. number of common species among two samples, b = number of species of one sample and c = number of 2: Frtngtng reef at the northern face of Siriba Island species of the other sample compared) (Legendre & Legendre, 1979). Cluster analysis using UPGMA This profUe was 32 m longo The substratum is strategy was performed by FITOPAC (Sheperd, formed by boulders and sand in the first 10 m (0.0 to 1994) software package. 0.8 m depth), with a low density of sessile benthic The species abundance and. diversity were also organisms. The fOllowing 13 m (1.0 to 1.6 m depth) calculated for coraIs but only to compare it with other are marked by dense ITondose macroalgae growing studies (e.g. Liddell & Ohlhorst, 1987). on a carbonl\te sand substratum. The last 10 m (1.6 to 2.7 m depth) are characterized by irregular calcareous coralline substratum, massive and Results encrusting corais, occasional colonies of the gorgonaC@!UlP/($xaurella regia, turf algae and Qualitative description crustose cOflllline algae. The 10m transects were placed in the algae and in the coral zones. The five surveyed locations are qualitatively described below: 3: MushrQom-s.hapedreef at the southern border of the Timbebas r(Jf!fcomplex 1: Fringing reef at the southern face of Santa Barbara Island The top of this reef has approximately 40 m in diameter and can be subdivided into 3 distinct The profile length was 60 m ranging ITomOto zones: the central fiat depression, the crest, and the 4 m depth. In the first 20 m (0.0 to 0.8 m depth) the border. The central fiat, with approximately 4 m bottom is mainly covered by crustose coralline algae, deep, is characterized by high coral cover with turf algae and zoanthids. The next 25 m (0.8 to 2.5 crustose and turf algae growing between coral m depth) are characterized by Palythoa caribaeorum colonies. The crest reaches the low tide leveI and is and turf algae. The last 15 m (2 to 4 m depth) are mainly epvered by crustose coralline algae, vermetid distinguished by the presence of massive coraIs and gastropoos with the zoanthid Zoanthus spp filling in turf algae. The lower leveI of the profile is covered by some crevices. The border zone is covered by the ITondose algae sparsely distributed (mostly by ramos!\!hydrocoral Millepora alcicornis and some massive coraIs. Downwards the reef top there is a 38 Rev. bras. oceanogr., 45(112), 1997 very irregular wali profile. The substrate is occupied they were only present in 56 and 50% of the samples, by encrusting invertebrates (mainly sponges and respectively. ascidians), turf and coraliine algae, anthipatarians (Anthipates sp) and gorgonians. The transect was Table 2. List of the observed taxa in the line transect; the placed on the top of the reef in the central flat functional group (EG) proposed for each taxon on depression. this work(corals= Cor; zoanthids= Zoa; turf-= Tur, crustose corallinaeae= Cru; fi'ondose= Fro), relative fi'equency(Fr), taxonomic dominance (Dt) 4: Mushroom shaped reei at t~ Parcel dos Abrolhos and importance lndex I. Reefs Species EG Fr D, This reef has a columnar shape of SCLERACTINIA approximately 18 m in diameter and 12 m in height. Agaricia agaricites Cor 62 0.4 0.25 The top of the reef lies at approximately 6 m depth Favia gravida Cor 56 0.4 0.22 and it is distinguished by the presence of massive Mussismilia braziliensis Cor 56 8.9 4.98 brain coraIs. In addition, there are turf and crustose Montastraea cavernosa Cor 18 0.4 0.07 coraliine algae. Below the bord@f,there is a wali Mllssismilia hartti Cor 12 0.5 0.06 similar to the previously described reef, where Mllssismilia hispida Cor 18 0.5 0.09 different species of coraIs are seen with a very 10W PQrites astreoides Cor 12 0.4 0.05 rate of cover. Crustose coralline algae and turf algae Porites branneri Cor 18 0.1 0.02 were present along with other invertebrates like Siderastrea stellata Cor 62 1.7 1.05 sponges and ascidians. The transect was placed on MLILLEPORlNA the top of the reef. Mil/epora alcicornis Cor 19 1.3 0.25 ZOANTHIDEA 5: Mushroom shaped reei c/ose to Sueste Island Palythoa caribaeorum Zoa 50 12.3 6.15 Zoa 50 1.7 0.85 The reef top had about 10 m in diameter, and Zoanthlls sociathus 6 was 9 m deep. It was mainly covered by turf algae PORIFERA with a moderate cover of massive and encrusting CYANOPHYCEAE Tur 56 17.4 9.74 coraIs and a few colonies of the gorgonacea CHLOROPHYCEAE Phylogorgia dilatata. The walls were mainly covered Bryopsis pennata Tur 6 0.1 0.01 by turf algae with very low coral coverage. The Caulerpa spp Tur 12 1.7 0.20 transect was placed on the top ofthe reef. Udotea cyathiformis Fro 31 0.3 0.09 Ali studied reefs were surrounded by fine sand PHAEOPHYCEAE flats where patches of seagrass and macroalgae beds Dictyota cervicornis Fro 25 1.5 0.38 can be found. [);ctyota mertensii Fro 12 0.8 0.10 Dir;tyota sp Fro 25 0.5 0.13 Quantitative community analYllis Diçtypteris jllstii Fro 18 2.6 0.47 Dlr;tyopteris plagiogramma Fro 12 1.5 0.18 A total number of 34 taxa were observed in the Lobophora variegata Fro 19 0.1 0.02 line transects, most of all attributed to coraIs Padina santae-crucis Fro 37 0.4 0.15 (Scleractinia and Milleporina), 50ft coraIs Sargassum spp Fro 25 7.3 1.83 (Zoanthidea) and algae (Table 2). The substrata were Stypopodillm zonale Fro 31 1.9 0.59 completely covered by living organisms and the RHOOOPHYCEAE cover relation between coral and algae varies among Amphlroa sp Tur 6 0.2 0.01 the studied zones. As a general pattem, algae were Ceramia1es Tur 43 3.2 1.38 the organism with dominant percentage cover. Table Corallinaceae Cru 94 14.7 13.82 2 shows the relative frequency, the total dominance Gellidlllm pusillum Tur 93 14.4 13.39 and the index of importance (I) for each taxon. Fro 6 < 0.1 O Crustose coralline algae (Crustose group) and Hypnea çervicornis Jania adhaerens Tur 37 2.5 0.93 Gellidium pusillum (Turf group) were the most Cru 12 0.2 0.02 important species in ali studied sites. Furthermore, Neogoniollthon sp Fro 6 <0.1 O they were present in almost ali surveyed transects. Qchtodes secundiramea Cyanophyceae was another class of algae with a high index of importance. The most important cnidarian In order to have a better comparison of species in the studied communities were Mussismilia community structure from different sites, the taxa braziliensis and Palythoa caribaeorum, although were classified in 5 functional groups: hermatipic \Tll.,LAÇA & PITOMBO: Benthic communities of Abrolhos 39 corais, Zoanthidae, crustose coralline algae, fiondose in the community structure. AIgae (mainly turf algae and turf algae [the algal groups approaching algae) had 60% cover in the first year (Dec. 92) and the morphofunctional classification of Littler et ai. 66% in the following year (Nov. 93); coral cover was (1983) and Steneck & Dethier (1994)]. The 39.4% and 33.4%, respectively, for the same years dominance of each group in the surveyed transects is (Fig. 2). ShOWIlin Figure 2. Turf algae were the most Santa Barbara and Siriba Island ftinging reefs important group (in % cover) in the two coral zones were surveyed during two consecutive years. The of the studied fiinging reefs. In the mushroom-like coral zone, in both reefs, show negligible differences reefs turf algae were also the dominant group in conceming alga/coral dominance, except for the cover, except for Timbebas Reef, where coraIs were samples taken in December 1992 (Fig. 2). These dominant, although coraIs are the main reef builders differences were probably due to a measurement in the region along with crustose coralline algae. error, because it was not possible to locate the Coralline algae were not a dominant group in the transect line in the same exact place as the previous studied sites, but they were one of the most ftequent year. However, important differences between (Table 2). Despite the lower percent cover, dead coralline and turf algae cover were found in the coralline crusts are a very important reef substratum Santa Barbara reef (Fig. 2). The percentage cover of in this region. Zoanthids were a conspicuous turf algae in Santa Barbara reef for example, attained component of the shallow zone of the Santa Barbara more than 80% in percent cover in the first ftinging reef (Palythoa zone), and were also present measurement, around 50% in the second and in the in other areas, especiallyat Timbebas reef (Fig. 2). third sample it was close to 65%. This kind of Frondose algae were mostly restricted to the variation are possibly due to the dynamic of the fast algae zone (Fig. 2) where they presented a very dense growing filamentous algae which live under heavy cover. This pattem could be observed in some other herbivorous pressure, and eventually, they might reefs of the same region. have their biomass completely grazed, a common Three reefs were surveyed more than once, feature observed in others reefs (Hatcher & Larkum, always during summertime. In the Parcel dos 1983). Abrolhos Reef, no expressive changes were observed

100%

80%

60% I I _COR COVER ~CRU c=::::J ZOA IIIJ] l' 1<0 40% t::::J '1'IrR

20%

0% N3 N3 D2 N3 J2 D2 N3 Tirnbebas Sueste Parcel Siriba I. reef reef Abrolhos algae zone TRANSECTS

Fig. 2. Percentage of functional group dominance, ITomthe 16 transects: hermatipic corais (Cor), crustose coralline algae (Cru), zoanthids (Zoa) , ITondosealgae (Fro) and turfalgae (Tur). 40 Rev. bras. oceanogr., 45(112), 1997

The Palythoa zone at Santa Barbara Island also meters (Table 1). This difference in depth may showed annual differences both in algae and intluence the capacity of some coraIs species to zoanthids percent cover: 64.2%, 50.2% and 60.8% colonize such shallower environments, consequently for Zoanthids (Palythoa caribaeorum) and 37.6%, reducing their diversity. 49.8% and 39.2% for algae, respectively, in the last The cluster analysis on Sorensen similarity three years. It seems that due to bottom irregularity, index for all surveyed transects is presented in Figure very distinct in this zone, it was difficult to repeat 3. At 0.5 leveI it is possible to distinguish 3 major saropling exactly at the saroe previous studied site. groups: the fust one encloses the transects made in The algae zone at Siriba Island showed very the mushroom-shaped reefs and in the coral zones of dense cover of fi'ondose macroalgae, mainly the Santa Barbara and Siriba fi'inging reefs; the Phaeophyta. However, in the second surveyed year, second group is formed by the transects surveyed at turf algae, especially Gellidium pusillum, accounted the Palythoa zone of the Santa Barbara fi'inging reef, for 26.4% in cover versus 0.7 and 3.05%, and the third group corresponds to the algal zone respectively, one year before and one year after (Fig. transects at Siriba Island. At the 0.6 leveI the fust 2). major group is subdivided in two: one group The reef diversity indices are shown in Table 3. corresponding to the mushroom-shaped reefs, which In general, the diversity is higher for the mushroom are segregated fi'om a second group belonging to the shape reefs and lower for the Santa Barbara Island fi'inging reef coral zones, except for one transect reef. Care should be taken in regard to these data, as placed in a coral zone of Siriba Island. the turf algal diversity was underestimated. A detailed laboratOry observation of a turf fi'om the coral zone of Siriba reef revealed the presence of 13 taxa. In the field, it was only possible to identify the Discussion major components of the turf. !tis assumed that the saroe error was introduced in all transects surveyed. Many studies of coral reef community structure fail to present data on the algal diversity and Table 3. Number of species (N); species diversitj (H') and coverage, concentrating instead on coraIs (eg. Loya, Pielou eveness (E) for all taxonomic units, in each 1972; Cantera, 1983; Huston, 1985; Pitombo et ai., transect. Coral species number (CN) species diversity (cH') and eveness (cE) is also presented 1988). Loya (op. cit.) reported high coral cover on different reef zones in the Red Sea ranging fi'om 22% Transect N H' E cN cH' cE for depths between 13 to 19 m, to 88% at 30 m. BCJ2 9 1.22 0.39 2 0.15 0.54 BCD2 10 2.07 0.62 3 0.60 0.22 Cantera (op. cit.) found that coral cover vary fi'om BCN3 10 1.79 0.54 3 0.22 0.20 almost zero in the boat channel to 80% on the fore BPJ2 7 2.06 0.74 I I I reef slope on a Colombian fi'inging reef, but he gave BPD2 6 1.93 0.75 1 I I BPN3 6 1.70 0.66 I I I no information about algal cover. Huston (op. cit.) SAJ2 11 2.32 0.67 I I I found 8 to 69% of coral cover in Discovery Bay, SAD2 11 2.32 0.72 I I I Jaroaica. Pitombo et ai. (op. cit.), found a maximum SAN3 9 2.39 0.75 I I I SCJ2 12 2.39 0.67 5 0.86 0.53 of 30% of coral cover at Siriba fi'inging reef, in SCD2 14 2.19 0.58 5 0.95 0.59 Abrolhos. In this study some functional groups other SCN3 14 2.88 0.76 5 0.4 0.25 than coraIs were assessed. Coral together with ROD2 10 2.96 0.89 3 0.72 0.66 crustose corallinacea are the main reef builders of the RON3 12 2.96 0.82 5 1.02 0.63 SUN3 15 2.60 0.66 6 1.31 0.73 Abrolhos system (see Fig. 2) and the contribution of TIN3 14 2.85 0.75 6 0.92 0.51 crustose corallinacea to the bulk of the reef must be evaluated in different reefs in order to have a better In the Siriba reef algal zone, where fi'ondose understand of this system. macroalgae was the dominant cover, and the Data fi'om the present study suggest that underwater identification was easier to perform, the mushroom-reef tops are qualitatively very similar to species diversity was very low in comparison with a the coral zones of fi'inging reefs (Fig. 3). Two.other typical tropical marine tlora (Yoneshigue, 1985). studied communities, zoanthid and algal zones, have The reef diversity using only the coral species, was distinctive compositions and must be analyzed very low in general but was somewhat higher in the separately. Furthermore, line transects do not seem to. mushroom-shaped reefs (Table 3). Both studied reefs be the appropriate method for studying fi'ondose were very shallow not attaining depths below 4 algal beds as they have a mobile and tridimensional meters. On the other hand the mushroom-like reef structure. When quantitative data are included in the transects were positioned in depths always below 4 Vll.LAÇA & PITOMBO: Benthic communities of Abrolhos 41

0.2

0.3

0.4

:.>t E-t 0.5

; 0.6 I {/'J I I 0.7 I

,--'--- 0.8 -'--- ,--'---

0.9 -'---

1..0 - - T R R S S B B S S B B B S S S [ D D U C p P P A A A N N D J N J N D N D N J D N J 3 3 2 2 3 2 3 2 3 2 3 2 2 3 2

TRANSECTS

Fig. 3. Cluster analysis ofthe 16 line transects using UPGMA strategy and SBrensen similarity index, the transects codes are furnished in Table 1. - analysis, some differences appear between the top of Therefore, it would be expected that these mushroom-reefs and the coral zones of fringing cQmmunities evolved a great stability. reefs. The turf group is more important in the cover The shallow communities of the Abrolhos reefs ofthe coral zones (Fig. 2). are clearly dominated by algae, mainly filamentous Shannan diversity was higher in mushroom turf algae. This pattem is also known in many other Teefswith or without algal components (Table 3). For reefs (Hatcher & Larkum, 1983; Klumpp & coral species it is c1ear that besides Mussismi/ia McKinnon, 1992). Besides that, most of the braziliensis, no other hard coral was important in the filamentous algal cover is mainly composed of cover of fringing reefs. These data agree with the Cyanophyceae, suggesting a system with great observations of Pitombo et aI. (1988). The difference capacity for fixing nitrogen (Wiebe et aI., 1975). The in diversity found among mushroom-like reefs and major current in the Abrolhos region is the Brazilian fringing reefs may be due to reef depths. The Current, which is poor in nutrients. This may also fringing reefs are very shallow, not attaining depths contribute to the development of the fast growing higher than 4 meters. In this case, the main nitrogen fixing Cyanophyceae which would be an constructor coral species cannot develop well, so advantage in pour environment. In Paredes Reef, for there is less spatial heterogeneity. In the other hand, example, frondose fleshy macroalgae are very mushroom-like reef transects were laid in depths abundant in cover and biomass (Coutinho et aI, below 4 m (Table 1), where very complex structures 1993). This reef is localised c10se to the coast in are formed and thus supporting a higher diversity. shallower water, receiving more nutrients inputs The three major studied zones (coral, Palythoa ftom thecoastal ecosystems (eg. mangroves and and alga) did not show any marked annual villages) and by sedimentsressuspension due to differences. The studied reefs are not affected by wind driven forces. storm perturbations. Disturbances caused by Hatcher (1990) hypothesized that many factors hurricanes, that are common to many other reef such as herbivory, local nutrient concentration, algae systems, are unknown in the Abrolhos area. defenses effectiveness etc... affect plant biomass on 42 Rev. bras. oceanogr., 45(1/2), 1997 coral reefs. Littler & Littler (1984) stated that Hatcher, B.G. 1990. Coral reef productivity: A nutrient levels and herbivore activity act as two hierarchy of pattem and processo Trends Ecol. major factors controlling the dominance of benthic Evolution,5:149-155. primary producers: coraIs, microfilamentous algae, frondose macroalgae and coralline algae. Herbivory Hatcher, B. G. & Larkum, A. W. D. 1983. An is a common biological factor in structuring reef experimental analysis of factors controlling the communities all over the world (review in Steneck, standing crop of the epilithic algal community 1988). In the studied reefs, fishes are usually the only on a coral reef. 1. expl mar. Biol. Ecol., 69:61- population of herbivores which can exert high 84. herbivore pressure. Sea urchins are not very common, even during the night (personal Huston, M. 1985. Pattems of species diversity in observation). When present, urchins usually stay in relation to depth at Discovery Bay, Jamaica. protected crevices of dead coralline algae bottoms. Bull. mar. Sei., 37:928-935. In the present study no specific experiment in herbivory was performed but some speculation could Klumpp, D. W. & McKinnon, A. D. 1992. be assumed in the light of the functional group Community structure, biomass and productivity approach (Littler et al., 1983). Considering the of epilithic algal communities on the Great relation dominance paradigm of Littler and Littler Barrier Reef: dynamics at different spatial (op. cit.), the studied reefs are characterized by low scales. Mar. Ecol. Prog. Ser., 86:77-89. nutrient levels and low herbivore activity (dominance of filamentous turf algae). The areas of coral Leão, Z. M. A. N; Araújo, T. N. F. & Nolasco, M. C. dominance (eg. Timbebas reefs) or patches of 1988. The coral reefs off the coast of eastem crustose coralline (coral zone of fringing reefs) might Brazil. In: Choat, J. H. et aI. eds Proceedings of be under pressure ofheavy herbivore activity. the 6th Intemational Coral Reef Symposium, 3:339-347.

Acknowledgments Leão, Z. M. A. N; Bittencourt, A. C. S. P; We are grateful to the staff of the Parque Dominguez, 1. M. L; Nolasco M. C. & Martin, Nacional Marinho dos Abrolhos (MNA-IBAMA) L. 1985. The effects of Holocene sea leveI that gave us funds and field facilities. fluctuations on the morphology of the Brazilian coral reefs. Rev. bras. Geoc., 15:154-157.

References Laborel, J. 1969a. Madréporaires et Hydrocoraliaires récifaux des cotes Brésiliennes. Systématique, Cantera, J. R 1983. Distribution des peuplements de écologie, répartition verticale et géographique. scléractiniaires sur un récif frangeant de l'íle de Annls Inst. oceanogr ., 47:171-229. Gorgona (cote Pacifique de Colombie). Téthys, 11:24-31. Laborel, J.1969b. Les peuplements de madreporaires des cotes tropicales du Brésil. Ann Univ Coutinho, R; Villaça, R; Magalhães, C. A; Abidjan, ser. E, Ecologie, 2:1-261. Guimaraens, M. A.; Apolinario, M. & Murici, G. 1993. Influência antrópica nos ecossistemas Legendre, L.& Legendre, P. 1979. Ecologie coralinos da região da Abrolhos, Bahia, Brasil. numerique.Tomme2. La structuredes donnes Acta Biol. Leopoldensia, 15:133-144. ecologiques.Paris,Mason.254p.

Gonchorosky, 1.; Sales, G.; Belém, M. J. C. & Liddell, W. D. & Ohlhorst, S. L. 1987. Pattems of Castro, C. B. 1989. Importance, establishment reef community structure, north Jamaica. Bull. mar. Sci., 40:311-329. and management plan of the "Parque Nacional Marinho dos Abrolhos", Brazil. In: Neves, C. ed. Coastlines of Brazil. Part of a series of volumes on Coastlines ofthe World. New York, Littler, M. M.; Littler, D. S. & Taylor, P. R 1983. American Society of Civil Engineers. p.185- Evolutionary strategies in a tropical barrier reef 194. system: functional-form groups of marine macroalgae. 1. Phycol., 19:229-237. VILLAÇA& PITOMBO: Benthic communities of Abrolhos 43

Littler, M. M. & Littler, D. S. 1984. Models of Steneck, R. S. & Dethier, M. N. 1994. A functional tropical reef biogenesis: The contribution of group approach to the structure of algal- algae. In: Round, F. E. & Chapman, D. J. eds dominated communities. Oikos, 69:476-498 Progress in Phycological Research. Amsterdam, EIsevier Biomedical Press. v. 3, p. 323-364. Wiebe, W. J.; Johanes, R. E. & Webb, K. L. 1975. Nitrogen fixation in a coral reef community. Loya, Y. 1972. Communitystructure and species Science, 188:257-259. diversityofhermatypiccoraIsat Eilat, Red Sea. Mar.Biol. 13:100-123. Yoneshigue,Y. 1985. Taxonomie et ecologie des algues marines dans Ia region de Cabo Pitombo, F. B.; Ratto, C. C. & Belém, M. J. C. 1988. Frio (Rio de Janeiro, Bresil). Ph.D. Thesis. Species diversity and zonation pattern of Université d'Aix-Marseille, Faculte des hermatypic coraIs at two ftinging reefs of Sciencesde Luminy.466p. Abrolhos Archipelago, Brazil. In: Choat, J. H. et ai. eds. Proceedings of the 6th International Coral ReefSymposium, 2:817-820.

Sheperd, G. J. 1994. FITOPAC 1. Manual do usuário. Universidade Estadual de Campinas, Depto de Botânica.

Steneck, R S. 1988. Herbivory on coral reefs: a synthesis. In: Choat, J. H. et ai. eds. Proceedings of the 6th International Coral Reef (Manuscript received 11 Apri11997; revised Symposium,I:37-49. 12 February 1997; accepted 12 March 1997)